Linkage disequilibria among closely linked restriction fragment length polymorphisms (RFLPs) in 50 genomic regions will be evaluated in order to answer two questions: (1) How much additional information for linkage studies is gained by using multiple closely linked RFLPs (compound loci) as markers instead of single RFLPs? The effect of linkage disequilibrium on reducing the heterozygosity of each compound locus will be measured. If linkage disequilibrium is high among RFLPs within compound loci, then multiple closely linked RFLPs contribute little added information. (2) Is there a uniform relationship between linkage disequilibrium and the physical distance between closely linked RFLPs? Linkage disequilibrium values have recently been used to infer gene order. This assumes a uniform inverse relationship between disequilibrium and the recombination fraction between loci. It is not known whether this relationship holds in very small genomic regions, where other forces, such as mutation, gene conversion, and drift, may play a large role. Few previous tests of this relationship have been done, and their results have conflicted. The statistical comparison of physical distance and linkage disequilibrium proposed here will be carried out on a large sample of genomic regions from a large number of individuals in the same, well-defined population. In addition to assessing standard linkage disequilibrium, 3- and 4-locus measures will be used to test for higher-order interactions. When possible, mechanisms that might cause variation in recombination rates between closely spaced RFLPs, such as tandem repeats or chi-like sequences, will be evaluated. These results will have important implications for the mapping of disease genes.